QuickStart

Start once detection within 5 lines of code.

from streamad.util import StreamGenerator, UnivariateDS
from streamad.model import SpotDetector

ds = UnivariateDS()
stream = StreamGenerator(ds.data)
detector = SpotDetector()

for x in stream.iter_item():
    score = detector.fit_score(x)
    # print("\r Anomaly score: {}".format(score), end="", flush="True")

Univariate time series

Detection, evaluation and visualization. Note that the gray line is the ground truth of anomalies.

from streamad.util import StreamGenerator, UnivariateDS, AUCMetric, plot
from streamad.model import SpotDetector

ds = UnivariateDS()
stream = StreamGenerator(ds.data)
model = SpotDetector()

scores = []

for x in stream.iter_item():
    score = model.fit_score(x)
    scores.append(score)
    # print("\r Anomaly score: {}".format(score), end="", flush="True")

data, label, date, features = ds.data, ds.label, ds.date, ds.features
auc_score = AUCMetric().evaluate(label, scores)
plot(data,label,date,scores,features)

Multivariate time series

An example of multivariate time series detection.

from streamad.util import StreamGenerator, MultivariateDS, AUCMetric, plot
from streamad.model import xStreamDetector

ds = MultivariateDS()
stream = StreamGenerator(ds.data)
model = xStreamDetector()

scores = []

for x in stream.iter_item():
    score = model.fit_score(x)
    scores.append(score)
    # print("\r Anomaly score: {}".format(score), end="", flush="True")

data, label, date, features = ds.data, ds.label, ds.date, ds.features
auc_score = AUCMetric().evaluate(label, scores)
plot(data,label,date,scores,features)